Table 3 CKD binary classification metric results presented as means of the 10 folds plus–minus the standard deviation.

From: Machine learning for classifying chronic kidney disease and predicting creatinine levels using at-home measurements

 

Features

Entropy

Accuracy

TPR

TNR

FPR

FNR

ANN

At-home

\(0.349 \pm 0.116\)

\(82.9 \pm 9.93\)

\(92.0 \pm 7.00\)

\(67.9 \pm 34.8\)

\(32.1 \pm 34.8\)

\(8.03 \pm 7.00\)

Monitoring

\(0.0800 \pm 0.0401\)

\(98.7 \pm 2.12\)

\(98.8 \pm 2.63\)

\(98.5 \pm 2.95\)

\(1.46 \pm 2.95\)

\(1.22 \pm 2.63\)

Laboratory

\(0.0404 \pm 0.0541\)

\(99.2 \pm 1.69\)

\(98.8 \pm 2.50\)

\(100.0 \pm 0.00\)

\(0.00 \pm 0.00\)

\(1.19 \pm 2.50\)

RF

At-home

\(0.215 \pm 0.0680\)

\(92.5 \pm 4.08\)

\(90.0 \pm 6.63\)

\(95.8 \pm 5.50\)

\(4.23 \pm 5.50\)

\(9.96 \pm 6.63\)

Monitoring

\(0.0714 \pm 0.0315\)

\(98.7 \pm 1.77\)

\(99.0 \pm 1.97\)

\(98.0 \pm 3.08\)

\(1.96 \pm 3.08\)

\(0.972 \pm 1.97\)

Laboratory

\(0.0394 \pm 0.0199\)

\(99.5 \pm 1.05\)

\(99.6 \pm 1.25\)

\(99.2 \pm 2.50\)

\(0.833 \pm 2.50\)

\(0.417 \pm 1.25\)